First Things First: The Hypothesis
Everyone who engages in the pursuit of new knowledge through research is aware of
the “Ask a question-perform a literature review-construct a hypothesis” cascade. However,
this process is not always rigorously followed in grants development. This is a simple
primer to remind investigators of the value of deliberate adherence to this cascade.
Ask a Question: The initial question is normally driven by background, interests, capabilities, and institutional core resources. The question may be a logical extension of research engaged in the past. Avoid questions that are strictly driven by funding opportunities if these are not questions you are passionate about or do not tap into your competences. However, you are encouraged to review the relevant funding agencies’ publications and websites to see if your question fits within their priority areas. This review of available funding may also be used as the first opportunity to refine your question.
Perform a Literature Review: A literature review should summarize and synthesize what is known about your question. It should also identify gaps in our knowledge and points of controversy. Hopefully, your initial question corresponds to one of the latter categories. Avoid the temptation to simply generate a list of findings; seek to organize findings from the literature into themes or concepts. An electronic reference manager, such as Endnote, helps tremendously. Revisit your question in light of the completed literature review.
Construct the Hypothesis: With limited exceptions, a competitive research application must be based on a single focused and testable hypothesis that predicts an outcome that increases the understanding of a biological process, treatment, or disease prevention. The NIAID publication, “How to Plan a Grant Application,” suggests that, if you have more than one hypothesis, choose the better one. The hypothesis should clearly tie to previous studies cited in background and your own preliminary studies. Each of your specific aims must, in turn, test some aspect of the hypothesis.